Risc OS on Raspberry Pi: Still having key repeat issues!

Hi there!

I’m a very happy Risc OS user on the Raspberry Pi, mostly thanks to the awesome piece of software that ADFFS is and Jon Abbott’s JASP project in general, which is something that brings back a lot of fun and tinkering to my computer life :D

However, I am havig stuck key issues since I started using Risc OS on the Pi. It happens when I try to code something, when I write commands on the F12 console…
I am using a wireless Ligitech keyboard which is directly connected to the Pi, no HUBs, nothing. My power supply is 2A, so that’s not the problem.
AND IT NEVER, EVER HAPPENS IN RASPBIAN, which makes the issue specially irritating, because I like Risc OS as a desktop OS WAY more than GNU/Linux.
I am using RC14 ROM, as newer versions seems to break ADFFS compatibility, but I guess the issue is still there, so even if I update, have changes been introduced to the USB code since RC14 that could cure this issue?
We really need something as stable and compatible as RC14, but with the key repeat problems fixed at last.

My power supply is 2A, so that’s not the problem.

In fact, it is. Even 20A will not change the fact that the Pi gives very limited power on USB ports. A powered hub is the only solution.
I suspect that Linux has an internal patch/trick for this.

I love ADFFS too :)

But then, couldn’t Risc OS implement the same patch/trick? If it works for Linux it should work for Risc OS too.

If you are feeling brave, you could try a newer ROM and/or Pi firmware.

It would be interesting to see if you have the same issues with a different (wired) keyboard too.

Richard: I suppose I would need a pre-July 5 ROM, because July 5 one does that zero-page modification that breaks compatibility with a lot of software, including ADFFS. So where can I download older daily builds?

I don’t think upgrading to a new ROM will help you – there haven’t been any significant changes to the USB driver since RC14 was produced.

The version of the USB driver that RISC OS currently uses is based on this version of the Linux driver from April 2014. The Linux driver has been changed quite a bit since then (it’s now using the fiq_fsm USB driver), and now that that’s stabilised one of the things on my todo list is to try updating RISC OS to use it.

I’m not working on any other high-priority stuff at the moment, so I might be able to start work on it in the next few days.

That’s VERY good to hear, Jeffrey! I will be waiting to try new versions then :)

I suppose I would need a pre-July 5 ROM, because July 5 one does that zero-page modification that breaks compatibility with a lot of software, including ADFFS. So where can I download older daily builds?

As Jeffrey has said there aren’t a lot of applicable changes so the pre/post July 5th image is not likely to produce a better result.

However, if you ask nicely I keep an archive and you’re welcome to an image from any date in the archive.
I’m sat in the wifes business premises¹ at present and the archive is on the NAS at home which is accessible from the net otherwise I’d tell you the last image before July 5th.

¹ “Amethyst Centre” – alternate therapy stuff. Since none of you are near Coventry you probably aren’t interested

Could it be a good idea to put all of this on line?
(I can help)

the last image before July 5th.

The answer appears to be July 2nd

Could it be a good idea to put all of this on line?
(I can help)

I’d be inclined to ask ROOL before anyone starts making old images generally available. I’ve always passed on copies purely for bug testing purposes.

It would be interesting to see if you have the same issues with a different (wired) keyboard too.

On my RPi, I only get key repeat issues with a wireless keyboard. Wired keyboard seems to be fine.

SeñorNueces: have you tried a different keyboard?
The problem seems to affect some keyboards a lot more than others. The updates done in April last year fixed many/most? RISC OS users key repeating problems. I can’t recall seeing it since myself since.

I’d be inclined to ask ROOL before anyone starts making old images generally available. I’ve always passed on copies purely for bug testing purposes.

Yep normal.
Anyway, do you have the latest non-zero-page IOMD ROM? This one interests me :)

Anyway, do you have the latest non-zero-page IOMD ROM? This one interests me :)

It isn’t implemented on IOMD as yet, so that’s the one on the ROOL site.

To answer the overall question, IOMD, OMAP3, OMAP4, BCM2835 (RPi) and Tungsten(Iyonix) back to 2012 with some gaps where I was on holiday (it’s a manual process)

Oups :)

I’ve got mysef a PS/2 keyboard with an USB adaptor, and I tried to replicate the “stuck keys” problem. There’s no more stuck key problems with this one, but if I keep a key pressed for a while, it’s ignored after some seconds: for example, playing Twinworld, my character will stop by itself even if I have the “walk to the right” key pressed.
However, with another wireless USB keyboard, the stuck keys problem persists, while they never appear on Raspbian.

There’s no more stuck key problems with this one, but if I keep a key pressed for a while, it’s ignored after some seconds:

Possibly a silly question but does installing this make any difference to the keyboard behaviour?

@Steve: not a silly question at all! :) But no, it doesn’t make any difference.

Last night I checked in my port of the ‘FIQ FSM’ USB driver, so it should be available in today’s ROM. So if anyone who’s been having trouble with stuck keys could give it a go and report back with their findings then that would be great.

I’ve gone bleeding edge on my build Pi and will give it a good test. I was hoping it would fix the unresponsive keyboard under Wolfenstein 3D, but sadly not, so it’s more than likely a coding issue in the game itself.

Am I right in assuming USB no longer uses IRQ and is solely using FIQ? Just wondering what I should be enabling under the processor vectors to keep it responsive.

EDIT: Forgot to add that provided I don’t see any build related issues, I’ll put this ROM on the four Pi’s I’ll have in the games Arcade at the London show.

EDIT2: I’m still seeing stuck keys unfortunately…whist typing this on the Pi. I’ll swap out the PSU to double check.

As a matter of interest do you get stuck keys with my Pi rom

I’ve change the PSU to a brand new official 2A Pi PSU and am still seeing stuck keys. Typing this is proving a nightmare!

However, I’m using this keyboard through a Data Centre grade digital KVM at the minute so will also need to test a direct attached keyboard and a powered hub keyboard. It’s not made any difference in the past, but worth re-testing.

As a matter of interest do you get stuck keys with my Pi rom

Will test and let you know.

@Colin – looking good so far, haven’t seen a single stuck key. I would normally have seen a dozen or so by now.

I’ve retyped this message several times and tested a few games as well, looks good. Sadly, I have to go back to 5.23 and live with it whilst conducting ZeroPain testing.

Can your code/changes be merged into 5.23?

Am I right in assuming USB no longer uses IRQ and is solely using FIQ? Just wondering what I should be enabling under the processor vectors to keep it responsive.

Both IRQ and FIQ are used. The new driver moves more of the processing into the FIQ handler, but there’s still a lot of important stuff which happens under IRQ.

Excellent Jon. At least you know it isn’t a PSU problem and, as I haven’t merged Jeffrey’s usb changes yet, it isn’t an interrupt timing problem. I had to make changes to how interrupt endpoints work to get USB audio to work on the pi. I’m hoping Jeffrey’s changes will allow high res usb audio on the pi. At the moment it’s limited to 44100/16 (2 bytes per channel) before you start to get crackles.